•  
  •  
 

Article Title

STEP-COUNT ACCURACY OF PHYSICAL ACTIVITY MONITORS DURING PREGNANCY IN FREE-LIVING CONDITIONS

Abstract

Previous studies have assessed the validity and reliability of physical activity monitors worn by pregnant women under laboratory conditions. However, physical activity monitors have not been assessed under free-living pregnancy conditions. PURPOSE: 1) Determine the step-count accuracy of four commercially-available physical activity monitors worn by pregnant women under free-living conditions and 2) examine the effect of pregnancy trimester on monitor accuracy. METHODS: Participants were pregnant women (n=28) in their second or third trimesters who were 18-40 years of age and free of contraindications to exercise during pregnancy. Participants wore three consumer-grade activity monitors (FB, OM, NL) and two research-grade activity monitors (AG, SW) for three days of free-living activity during all waking hours. Steps recorded over the three days for the FB, OM, NL, and AG were compared to SW recorded steps (the criterion measure) in order to calculate percentage of actual steps taken ([measured steps / actual steps] x 100). Paired-samples t-tests were performed to determine differences in accuracy between monitors and one-way ANOVAs were utilized to determine whether pregnancy trimester affected monitor accuracy. RESULTS: The analytical sample consisted of 18 women in their second trimester and 10 women in their third trimester with an overall mean gestational age of 23.9 ± 8.19 weeks and a mean daily step-count of 9354.3 ± 3363.9 steps (as determined by SW). Steps taken per day did not significantly differ between second and third trimester women (F(1, 26)=0.69, p=0.42). The FB and NL were most accurate with mean percentage of actual steps taken recorded at 69.9% and 69.5% respectively. The AG (t(27)=-3.13, p<0.01) and OM (t(27)=-6.27, p<0.001) performed significantly worse with 62.7% and 52.1% of actual steps. Trimester did not significantly affect monitor accuracy. CONCLUSION: Compared to the criterion, all other monitors underestimated actual steps taken, with the FB and NL demonstrating smaller underestimations than the AG and OM in a free-living environment. Accuracy of these monitors appears to be worse during pregnancy free-living conditions compared to results of studies performed in controlled laboratory conditions.

Supported by WSU College of Education Faculty Funding Program

This document is currently not available here.

Share

COinS